An integrated model of tourists’ time use and expenditure behaviour with self-selection based on a fully nested Archimedean copula function

This study presents an integrated model of tourists’ time use, expenditure by explicitly incorporating the influence of destination visit behaviour based on the concept of self-selection. Inter-destination interactions (between time-use decisions, between expenditure decisions, and between decisions on both time use and expenditure) are incorporated by using a multilinear utility function. The study uses a Type-II Tobit model with a Frank-type fully nested Archimedean copula to deal with issues caused by correlation between these variables. The Archimedean copula is used to combine marginal distributions to obtain a joint distribution of error terms of utility functions with the help of dependency parameters. The effectiveness of the model is confirmed empirically by using data collected in the Tottori Prefecture of Japan in 2007. Dependence structures across destinations, inter-destination interactions and the relative importance of each destination in decisions are examined.